The overall objective of this research grant is to continue the development of modern statistical ideas for use in medical applications. This development involves the creation of new statistical methods, study of new and existing techniques through mathematical research and computer simulation, and trial application of these procedures to ongoing medical research projects at the Stanfdord Medical Center. The research will focus primarily on four specific areas: (1) forced randomization designs for clinical trials, (2) censored data analysis of survival times, (3) classification analysis for outcome prediction, and (4) robust/nonparametric methods of data analysis. The research will be conducted by the members of the Division of Biostatistics in the Stanford University School of Medicine. Application of the results of this statistical research to problems in cardiovascular research, cancer therapy clinical trials, and other areas is envisaged.